Vits-based Singing Voice Conversion Leveraging Whisper And Multi-scale F0 Modeling

Abstract

This paper introduces the T23 team's system submitted to the Singing Voice Conversion Challenge 2023. Following the recognition-synthesis framework, our singing conversion model is based on VITS, incorporating four key modules: a prior encoder, a posterior encoder, a decoder, and a parallel bank of transposed convolutions (PBTC) module. We particularly leverage Whisper, a powerful pre-trained ASR model, to extract bottleneck features (BNF) as the input of the prior encoder. Before BNF extraction, we perform pitch perturbation to the source signal to remove speaker timbre, which effectively avoids the leakage of the source speaker timbre to the target. Moreover, the PBTC module extracts multi-scale F0 as the auxiliary input to the prior encoder, thereby capturing better pitch variations of singing. We design a three-stage training strategy to better adapt the base model to the target speaker with limited target speaker data. Official challenge results show that our system has superior p

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyning2023vits

Related papers

• Vits-based Singing Voice Conversion System With DSPGAN Post-processing For SVCC2023 (2023)5.84
• A Comparative Study Of Voice Conversion Models With Large-scale Speech And Singing Data: The T13 Systems For The Singing Voice Conversion Challenge 2023 (2023)6.77
• Singing Voice Conversion With Non-parallel Data (2019)9.59
• Ppg-based Singing Voice Conversion With Adversarial Representation Learning (2020)9.76
• SYKI-SVC: Advancing Singing Voice Conversion With Post-processing Innovations And An Open-source Professional Testset (2025)4.52
• Real-time And Accurate: Zero-shot High-fidelity Singing Voice Conversion With Multi-condition Flow Synthesis (2024)0.00
• Sifisinger: A High-fidelity End-to-end Singing Voice Synthesizer Based On Source-filter Model (2024)4.52
• A Unified Model For Voice And Accent Conversion In Speech And Singing Using Self-supervised Learning And Feature Extraction (2024)0.00